The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
output_file argument to
tar_quarto() and tar_quarto_raw() for single
documents (#198, @mutlusun).tar_map()
and tar_map_rep() (#201, @kkmann).targets after
https://github.com/ropensci/targets/issues/1368.tar_quarto_inspect() (#200, @multusun).delimiter argument to tar_map() etc.
for customizing separators in target names (#177, @psychelzh).tar_assign() (#186,
https://github.com/ropensci/targets/issues/1309, @hadley).tar_format_feather() in favor of
tar_arrow_feather() (#190).tar_aws_*() target factories. They are
obsolete because of the repository argument in
tar_target().digest::digest() to
secretbase::siphash13().description argument of
tar_target() in targets 1.5.1.9001.tar_map() and other static branching target factories
now append values to the target descriptions. Use the
descriptions argument of those functions to customize.repository settings in
tar_change() and tar_map_rep().tar_knit(), tar_render(),
tar_quarto(), and their “rep” and “raw” versions all gain a
working_directory argument to change the working directory
the report knits from. Users who set working_directory need
to supply the store argument of tar_load() and
tar_read() relative to the working directory so the report
knows where to find the data (#169).tar_knit(), tar_render(),
tar_quarto(), and their “raw” versions all gain an
output_file argument to more conveniently set the file path
to the rendered output file.tar_quarto() and its “rep” and “raw” versions all gain
a new quarto_args argument for advanced Quarto arguments
(#166, @petrbouchal).tar_seed_create() and tar_seed_set()
from targets.tar_render() (#158).tar_rep2() compatible with tar_map()
(#165).tar_seed_create() and
tar_seed_set()
(https://github.com/ropensci/targets/issues/1139). Future versions of
tarchetypes should use these package functions, but this
version cannot because of the compatibility constraints of the release
cycle.targets >= 1.3.2.9004 progress
statuses (“completed” instead of “built”, “dispatched” instead of
“started”).packages and library
arguments of tar_quarto() and tar_quarto_raw()
(#155, @svraka).furrr to parallel for
rep_workers in tar_rep() etc. (#251, @solmos).targets
1.3.0.format = "file_fast" in target factories.QUARTO_PROFILE
environment variable (#139, @andrewheiss).targets issue 1047, @joelnitta).targets::tar_runtime_object()$store instead of
targets::tar_runtime_object()$get_store() to ensure forward
compatibility with targets.tar_download() to avoid
unpredictable network issues outside our control.set_deps argument in the hook functions
to force modified targets to keep the dependencies they had before
applying the hook (#131, @edalfon).tar_copy_target() (#131, @edalfon).tar_quarto_rep() and tar_render_rep() (#129,
@benzipperer).tar_quarto_rep() (#129,
@benzipperer).rep_workers to control inner parallelism in
batched replication functions (#117).furrr functions has
environment tar_option_get("envir").tar_quarto_rep() (#129, @mglev1n).targets::tar_path().tar_manfiest() default
output.tar_quarto() and
tar_quarto_rep() (and “raw” versions) for Quarto documents
and projects in pipelines (#89).tar_quarto_files() to inspect Quarto
projects and documents for important files: source files to inspect for
target dependencies, output documents, and Quarto-specific inputs like
_quarto.yml. Uses quarto::quarto_inspect() and
powers the automatic file detection in tar_quarto() etc.
(#89).params argument of
tar_render_rep() (must be a data frame with unique rows (or
unique elements of output_file)).root.dir when scanning for
dependencies so knitr child documents work (#93, @mutlusun).format = "rds" for target_batch in
tar_map_rep_raw() so the global format option does not mess
up the pipeline.tar_append_static_values().tar_render() to work with just one row of
parameters (#96, #97, @ugoebel73).memory and
garbage_collection arguments.tar_file_read() (#84, @petrbouchal).tar_select_targets()
(#92, @arcruz0).repository argument for
targets >= 0.11.0.command1 using
[[ and not [ in tar_map2()
functions.tar_map_rep() and
tar_map_rep_raw() for dynamic batched replication within
static branching for data frames (#78).tar_map2_count(),
tar_map2_count_raw(), tar_map2_size(), and
tar_map2_size_raw() for batched dynamic-within-static
branching for data frames (#78).tar_rep_map() in favor of
tar_rep2() to avoid name confusion. Likewise with
tar_rep_map_raw() to tar_rep2_raw()
(#78).NULL target list in
tar_map() (@kkami1115)."aws_file" format in
tar_files() or related target factories.targets timestamps correctly in the help files
of tar_age() and tar_cue_age().names = NULL in tar_map(), use hashes
instead of numeric indexes for generated target names (#67). That way,
target names are no longer sensitive to the order of
values, and so targets will incorrectly invalidate less
often. Unfortunately, this is an invalidating change: some targets
will automatically rerun after you install this version of
tarchetypes. I apologize for the inconvenience this
causes. However, we do need this patch in order to solve #67, and
targets will incorrectly invalidate less frequently in the future.targets (#59).*_raw() target factories process
command the same way whether it is an expression or
ordinary language object.targets 0.5.0.9000, which
logs skipped targets.tar_rep_map() and tar_rep_map_raw() to
perform batched computation downstream of tar_rep()
(#50).tar_select_names() and
tar_select_targets() to make certain metaprogramming tasks
easier.tar_map(), attempt to convert the elements of
values into lists of language objects.tar_plan() (#40, @kendonB).tar_age() based on tar_cue_age()
(#39, @petrbouchal).tar_cue_age(),
tar_cue_age_raw(), tar_cue_force(), and
tar_cue_skip() (#39).tar_download() (#38, @noamross, @petrbouchal)tar_render_rep() (#36, @gorgitko).tar_aws_rds() and
tar_parquet().tar_hook_before(),
tar_hook_inner(), and tar_hook_outer()
(#44).tar_map().crayon.enabled for literate programming.%||% and %|||% to
conform to historical precedent.tar_group_by(), tar_group_select(),
tar_group_size(), tar_group_count() (#32,
@liutiming).tar_render() and related functions, track the
*_files/ output directory if it exists (#30).walk_ast() function to make it
easier for other developers to extend the static code analysis of
tarchetypes (@MilesMcBain).targets::tar_dir(). (targets::tar_dir() and
targets::tar_test() already run code in a temporary
directory.)targets::tar_dir() runs code in a temporary directory,
which means all ostensibly files created in the enclosed expression will
actually be written to temporary storage and not the user’s file
space.man/ has
Rd-tags \value and \arguments.man/, describe
the return value in the \value Rd tag. For each function
that returns a target object or list of target objects, the
\value tag now links to https://books.ropensci.org/targets/, the user manual
where the purpose of target objects is explained, and https://books.ropensci.org/targets-design/, the design
specification which documents the structure and composition of target
objects.tarchetypes
package writes by default to the home file space of the user. The paths
of all output files are controlled by non-tarchetypes
functions that invoke tarchetypes.tar_plan() now returns a list of target objects rather
than a pipeline object. Related: https://github.com/ropensci/targets/issues/253.tar_knitr_deps() and
tar_knitr_deps_expr() to accommodate custom multi-file
literate programming projects like R Markdown sites and
bookdown projects (#23, @tjmahr).These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.